| { |
| "best_metric": 0.9296, |
| "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0617/checkpoints/checkpoint-1332", |
| "epoch": 4.0, |
| "eval_steps": 500, |
| "global_step": 1332, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 1.0, |
| "grad_norm": 1.6354416608810425, |
| "learning_rate": 0.00025606601717798207, |
| "loss": 0.6558, |
| "step": 333 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.876, |
| "eval_loss": 0.4066951274871826, |
| "eval_runtime": 6.4615, |
| "eval_samples_per_second": 580.359, |
| "eval_steps_per_second": 9.131, |
| "step": 333 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 4.341063499450684, |
| "learning_rate": 0.00015, |
| "loss": 0.2489, |
| "step": 666 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.9, |
| "eval_loss": 0.33929693698883057, |
| "eval_runtime": 7.012, |
| "eval_samples_per_second": 534.795, |
| "eval_steps_per_second": 8.414, |
| "step": 666 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 0.120743028819561, |
| "learning_rate": 4.3933982822017876e-05, |
| "loss": 0.0751, |
| "step": 999 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.9202666666666667, |
| "eval_loss": 0.28302833437919617, |
| "eval_runtime": 6.6561, |
| "eval_samples_per_second": 563.395, |
| "eval_steps_per_second": 8.864, |
| "step": 999 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 0.23546642065048218, |
| "learning_rate": 0.0, |
| "loss": 0.0162, |
| "step": 1332 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.9296, |
| "eval_loss": 0.2542307674884796, |
| "eval_runtime": 6.3878, |
| "eval_samples_per_second": 587.054, |
| "eval_steps_per_second": 9.236, |
| "step": 1332 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 1332, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 4, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 6.58965282490368e+18, |
| "train_batch_size": 64, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
